III - VIVEIROS DE CASTRO, Eduardo. 2003. (Anthropology) and (Science)
After-dinner speech at ‘Anthropology and Science’, the 5th Decennial Conference of the Association of Social Anthropologists of Great Britain and Commonwealth, 14 July 2003. Published in Manchester Papers in Social Anthropology, 7, 2003 Eduardo Viveiros de Castro --- Professor Fardon, distinguished colleagues, I’ve spent half my time over the last few months asking myself why the conference organizers decided to bestow the overwhelming honour of inviting me to address you all on this occasion. And the other half I’ve spent asking myself how I had the gall to accept such an invitation. Eventually I came to the conclusion that the answer to both questions is probably one and the same: both our hosts and I like to live dangerously. Indeed I have a sneaking suspicion Penny Harvey, Peter Wade and Jeanette Edwards were looking for the most unlikely person to speak at an ASA Conference dedicated to the theme of ‘anthropology and science:’ someone, let’s say, resembling an obscure foreign scholar who doesn’t practice an especially scientific anthropology, who has never undertaken any kind of anthropological study of science, and who, on top of all this, speaks English rather quirkily. I just hope they haven’t gone too far in their eagerness to surprise you. As for myself, suffice to say that the responsibility of succeeding the great Sahlins in marshalling your postprandial entertainment could only have been taken on by someone blessed with the most complete sense of irresponsibility. Even more so since Sahlins left you, ten years ago, waiting for none other than Foucault… Look what you’ve got instead — I’m not even bald. Still, I accepted the challenge of attempting to amuse you with some trifles on the theme ‘Anthropology and Science’ because of this little magic word ‘and’ — a connective which is to the universe of relations as the notion of mana (I mean Lévi-Straussian mana) is to the universe of substances. ‘And’ is a kind of zero-relator, a relational mana of sorts — the floating signifier of the class of connectives — whose function is to oppose the absence of relation, but without specifying any relation in particular. ‘And’ covers all thinkable connections, and therefore allows one to say all sayable things about the terms it connects — which naturally enough doesn’t demand the work of a specialist. Indeed this explains how I plucked up the courage to come here. But maybe not. Maybe there is a relation which ‘and’ excludes, perhaps because it is not a true relation — the relation of identity. Who would dream of giving a physics conference the title ‘Physics and Science’? Physics is Science! We have to be able to imagine that anthropology isn’t constitutively a science, at least not all the time, in all respects and in all relations, in order for us to imagine this contingent connection expressed in the formula ‘anthropology and science.’ A relation can be contrived, then, between ‘and,’ the minimal relator, and ‘is,’ the maximal substantializer, poles between which all our discourses and sciences are distributed. Now, if anthropology ‘is’ a science of something, it is undoubtedly the comparative science of the relations that make us human. But since comparing is relating, and vice-versa, our discipline is twice over the science of the ‘and,’ that is, of universal relational immanence. Not of the ‘is’, therefore, and still less of the ‘ought’ — but simply of the ‘and’. Everyone here will recall the famous last words of Primitive Culture: ours is announced as a reformer’s science, a ‘Ghostbusters’-like enterprise committed to tracking down and wiping out all superstition. Later on we learnt how to functionalize and rationalize superstition, arguing that it was merely an unselfconscious metaphoric sociology or an evolutionary spandrel precipitated by the cognitive make-up of the human species. Be that as it may, the fact that we have always defined anthropology, officially or officiously, as the science of non-science imbues the recent interest in an ‘anthropology of science’ with a reflexive piquancy all its own. The discomfort provoked by the idea of an anthropological description of scientific activity — a queasiness felt not just by practitioners of the hard sciences, but also by many anthropologists — suggests we are seen, and maybe we even see ourselves, as an accursed race of anti-Midases capable of transforming everything we touch into error, ideology, myth and illusion. So danger looms when the reformer’s science turns its gaze to science at large: the latter seems set to be denounced as just one more kind of superstition. This was how the so-called Science Wars, or Culture Wars, exploded, in which anthropologists featured among the prime suspects — based as usual on somewhat fabricated evidence — accused of possessing weapons of mass destruction. Or should I say, mass deconstruction. Of course it should be the complete opposite of all this. What the anthropology of science should be teaching us — and this, for myself at least, is its primary lesson — is the impossibility of continuing to practice our discipline within an economy of knowledge where the anthropological concept functions as a kind of surplus value extracted by the ‘observer’ from the existential labour — the life — of the ‘observed.’ What follows is an attempt to make this clearer. Obviously I cannot speak here for all my generation, those of us who turned adults around 1968, but for many of us anthropology was and still is the absolute opposite of a reformer’s science or a Reason police. It was an insurrectionary, subversive science; more specifically, the instrument of a certain revolutionary utopia which fought for the conceptual self-determination of all the planet’s minorities, a fight we saw as an indispensable accompaniment to their political self-determination. In the case of Brazilian anthropologists, this possessed an especially urgent relevance. The start of the 1970s saw the indigenous minorities in my country begin to establish themselves as political agents. Our aim as anthropologists was to assist this process by providing it with a radical intellectual dimension, enabling the thought of American peoples to escape the ghetto in which it had been enclosed since the 16th century. As part of this politico-cultural struggle, which may be imagined as a process of multiplicity-building (that is, of anti-empire building), the work of Lévi-Strauss — some of you may be surprised to hear, others not so — was of enormous importance, since it was through Lévi-Strauss’s mediation that the intellectual style of Amerindian societies was for the first time in a position to modify the terms of the anthropological debate as a whole. In sum, for us the expression ‘la pensée sauvage’ did not signify ‘the savage mind.’ To us it meant untamed thought, unsubdued thought, wild thought. Thought against the State, if you will. (In remembrance of Pierre Clastres). Sure enough, all of us were hippies of a kind. We were primitivists, anarchists and essentialists; perhaps we had a slightly inflated sense of anthropology’s importance; we were all highly prone to exoticism, too. But we weren’t quite so hopelessly naive: our primitivism was a desire for self-transformation; our anarchism needs no excuse; our essentialism was strategic (but of course); and as for our exoticism, well, those were strange times indeed when the concept of the Other designated a radically positive value, while the concept of Self was a position to be detested. In other words, our world had yet to wake up to the now pervasive sentiment against differ¬ence and alterity which sees them as harbingers of violence and oppression. All difference seems nowadays to be read as an opposition, while alterity is conceived as the absence of a relation: ‘to oppose’ is taken to be synonymous with ‘to exclude’ — a weird idea, which I can only put down to the guilty supposition that others conceive otherness as we do. Well, they don’t: others are ‘other’ precisely because they have other ‘others’ — Captain Cook, for example, as Sahlins has memorably argued. Anyway, I guess there’s no need to remind you that ‘othering’ is not the same kind of politico-metaphysical swindle everywhere. And come to think of it, why should ‘saming’ be such a better thing to do to others? Who wants to be samed? All those double-bind claims to ‘tolerance’ rush to mind; as the philosopher Isabelle Stengers asks — would you like to be tolerated? In my view, anthropology is consistently guided by this one cardinal value: working to create the conditions for the conceptual, I mean ontological, self-determination of people. Or peoples to be more exact. Its success or failure as a science hinges on this, and not, as some of our more nihilistic colleagues wishfully think, on its willingness to proclaim its own self-extinction and divide its legacy between a neo-evolutionist psychology and a neo-diffusionist history; yuppifying itself out of existence, in effect, not with a big bang but a spluttering whimper. In fact maybe it’s time for us to reinvent a neo-functionalist social anthropology…? Since we’re living through a moment in our discipline’s history when it seems increasingly urgent for us to reclaim and proclaim the very dimension of reality with which anthropology concerns itself: a collective reality — a relational reality, in other words —, one possessing a disposition towards the transcontextual stability of form. I believe anthropology must escape self-imposed doom and keep firmly focused on its proper object: social relations in all their variations. Not social relations taken as a distinct ontological domain (there is no such thing), but all phenomena as potentially comprising or implying social relations. This means taking all relations as social. Not though from a viewpoint completely dominated by the western doctrine of social relations, but from one ready and willing to admit that treating all relations as social may entail a radical reconceptualization of what ‘the social’ may be. Indeed anthropology distinguishes itself from other discourses on human sociality by not possessing a particularly solid doctrine on the nature of social relations. On the contrary, it tends from the outset to have only a vague idea of what a relation may be, since its problem typically consists not so much in determining what social relations constitute its object, but in asking itself what its object constitutes as a social relation. In other words, what a social relation is in its object’s terms, or better still, in the terms formulated by the relation between the ‘anthropologist’ and the ‘native.’ This of course leads us to our crunch question: what is an anthropologist, and who’s the native? The ‘anthropologist’ is someone who discourses on the discourse of a ‘native.’ The native need not be particularly savage, nor traditionalist, nor even natural to the place where the anthropologist finds him; the anthropologist need not be excessively civilized, nor modernist, nor even a stranger to the people about whom she discourses. The discourses of the anthropologist and above all the native are not necessarily texts: they are any kind of meaningful practice. The essential factor is that the discourse of the anthropologist (the ‘observer’) establishes a certain relationship with the discourse of the native (the ‘observed’). This relationship is a relation of meaning, or, as one says when the former discourse aspires to the status of a Science, a relation of knowledge. Such a relation is not one of identity: the anthropologist always says, and thus does, something different from the native, even if her intention is to do no more than ‘textually’ reiterate the native’s discourse, or contrive a ‘dialogue’ with him. Discursive alterity is of course premised on similarity. Anthropologist and native are entities of the same kind and condition: equally human and equally embedded within their respective cultures, which may even be one and the same. But it’s here that things start to become interesting, or should I say, strange. Even when the anthropologist and native share the same culture, the relation between the two discourses acts to differentiate this community: the anthropologist’s relation to her culture and that of the native to his are not exactly the same. What makes the native a native is the presupposition, on the part of the anthropologist, that the former’s relation to his culture is natural, that is, intrinsic and spontaneous, and, if possible, non-reflexive — or better still, unconscious. The native expresses his culture in his discourse; likewise the anthropologist, but if she intends to be something other than a native, she must express her culture culturally, that is, reflexively, conditionally and consciously. The anthropologist necessarily uses her culture; the native is sufficiently used by his. Needless to say, this difference isn’t to be found in the so-called nature of things. It’s an intrinsic element of the language game I’m describing and defines the figures labelled ‘the anthropologist’ and ‘the native.’ Let’s consider a few more rules of this game. The anthropological idea of culture places the anthropologist on equal terms with the native by implying that all anthropological knowledge of another culture is culturally mediated. However this equality is in the first instance purely empirical or de facto: it corresponds to the equally cultural condition of anthropologist and native. It doesn’t imply an equality de jure — an equality on the plane of knowledge. The anthropologist typically enjoys an epistemological advantage over the native. The two discourses are situated on different planes: the meaning established by the anthropologist depends on the native meaning, but it is she who determines this meaning’s meaning — she who explains and interprets, translates and relates, textualizes and contextualizes, justifies and signifies this meaning. The relational matrix of anthropological discourse is hylomorphic: the anthropologist’s meaning is form to the native’s matter. The native’s discourse can’t determine the meaning of its own meaning. As Geertz said somewhere, we are all (de facto) natives; sure; but some of us are (de jure) always more native than others. This prompts the following questions. What happens if we deny the anthropologist’s discourse its strategic advantage over the native’s discourse? What happens when the native’s discourse functions within the anthropologist’s discourse in such a way it produces a reciprocal ‘knowledge-effect’ on the latter? When the form intrinsic to the content of the first modifies the content implicit in the form of the second? Translator, traitor, as the Italian saying goes; but what happens if the translator decides to betray her own language? What would ensue if, dissatisfied with the mere passive or de fato equality between the subjects involved, we were to claim an active or de jure equality between the discourses themselves? In sum, what changes when anthropology is taken as a meaning-producing practice in epistemological continuity with the practices on which it discourses — as their equivalent? In other words, when we apply the Latourian notion of ‘symmetrical anthropology’ to anthropology itself, not to lambaste it as colonialist, exorcize its exoticism, or mine its intellectual field, but to induce it to say something completely different? Not only different from the native’s discourse, since this must remain one of anthropology’s functions, but different to the discourse which anthropology pronounces about itself, usually subvocally, when discoursing on the discourse of the native? Were we to pursue all this in active fashion, I would say that we would be doing what was always properly called ‘anthropology,’ instead of, say, ‘sociology’ or ‘psychology.’ I say I would say, because much of what was or is done in this name supposes, on the contrary, that the anthropologist holds total sway over those reasons of which the native’s reason knows nothing. She knows the exact doses of universality and particularity contained in the native, and the illusions which the latter entertains about himself — whether manifesting his native culture all the while believing he’s manifesting human nature (the native ideologizes without knowing), or manifesting human nature all the while believing he’s manifesting his native culture (he cognizes unawares). (Generally it’s supposed the native does both things without being aware of either — natural reasoning and cultural rationalizing — in different phases, registers or situations of his life. Moreover, the native’s illusions are taken as necessary in the double sense of inevitable and useful; they are, to hijack a phrase, evolutionarily adaptive. It is this necessity which defines the ‘native’ and distinguishes him from the ‘anthropologist:’ the latter may be wrong about the former, but the former must be deluded about himself.) Thus the anthropologist knows the native de jure, even though she may not know him de facto. The complete opposite occurs when moving from the native to the anthropologist: although he knows the anthropologist de facto (frequently better than she knows him), he doesn’t know her de jure, since the native is precisely not an anthropologist like the anthropologist. The anthropologist’s knowledge is a wholly different animal from the native’s knowledge. Indeed it has to be: the condition of possibility of the former entails the delegitimation of the claims of the latter, its ‘epistemocide’, in Bob Scholte’s forceful expression. Knowledge on the part of the subject requires a sort of transcendental nescience on the part of the object. It’s all very well — or rather, quite ill. But there is no reason for us to be excessively squeamish about all this. As the discipline’s history attests, this discursive game with its unequal rules has told us many an insightful thing about natives. Nevertheless, the experiment I propose here precisely involves refusing to play this game. Not because it produces objectively false results, or misrepresents the native’s nature, so to speak. Given the objects which the classic game takes as given, its results are very often convincing, or at least ‘plausible,’ as adepts of this game like to say. Refusing to play this game simply implies positing different objects compatible with different rules. What I’m suggesting in a nutshell is the need to choose between two conceptions of anthropology. On one side, we have an image of anthropological knowledge as the outcome of applying concepts extrinsic to its object: we know beforehand what social relations are, or cognition, kinship, religion, politics and so on, and our aim is to see how these entities take shape in this or that ethnographic context — how they take shape unbeknown to the interested parties, needless to say. On the other side (and this is the game I’m proposing), is an idea of anthropological knowledge which starts out from the premise that the procedures characterizing the investigation are conceptually of the same kind as those to be investigated. This equivalence at the level of procedures, we should note, supposes and produces a radical non-equivalence at all other levels. For while the first conception of anthropology imagines each culture or society as the embodiment of a specific solution to a generic problem — as the specification of a universal form (the anthropological concept) with a particular content (the indigenous representation) —, the second by contrast imagines that the problems themselves are radically distinct. More than this: it starts out from the principle that the anthropologist cannot know beforehand what these problems may be. Anthropology in this case places in relationship different problems, not a single (‘natural’) problem and its different (‘cultural’) solutions. Thus the ‘art of anthropology’ is to my mind the art of determining the problems posed by each culture, not the art of finding solutions to those problems posed by our own. This has been one of the most important lessons I’ve learnt from Marilyn Strathern. And it is for this very reason that the postulate of the continuity of procedures is an epistemological imperative. Of procedures, I repeat, not of those who carry them out. Since neither is it a question of condemning the classic game for producing subjectively falsified results by a failure to recognize the native’s condition as a Subject: by fixing him with a distant and cold gaze, constructing him as an exotic object, diminishing him as a primitive on another time-band to the observer, denying him the human right of interlocution — the litany is well known. Nothing of the sort, I believe. In fact very much the opposite: it is precisely because the anthropologist takes the native so readily as another subject that she fails to see him as an other subject, as a figure of Another who, prior to being a subject or object, is the expression of a possible world. It is by refusing to accept the native’s condition as a ‘non-subject’ (in the sense of being other than the subject) that the anthropologist introduces, under the guise of a proclaimed de facto equality with the former, her wily de jure advantage. She knows much too much about the native before the game even starts; she predefines and circumscribes the possible worlds expressed by this other; the alterity of the other is already radically separated from his capacity for alteration. The authentic animist is the anthropologist, and participant observation is the true (meaning false) primitive participation. Consequently the problem doesn’t reside in seeing the native as an object, nor does the solution reside in casting him as a subject. That the native is a subject is beyond doubt; but what the native forces the anthropologist to cast into doubt is precisely what a subject could be — such is the properly anthropological ‘cogitation.’ It alone allows anthropology to assume the virtual presence of Another as its condition, indeed precondition, and which determines the derivative and vicarious positions of subject and object. I evoked the Kantian distinction between quid facti and quid juris questions. It struck me as useful because the first problem to be solved involves evaluating the claim to knowledge implied in the anthropologist’s discourse. This problem is not cognitive or psychological; it doesn’t concern the empirical possibility of knowing another culture. It is epistemological — and thus political. It relates to the properly transcendental question of the legitimacy attributed to the discourses entering into a relation of knowledge, and, in particular, the relations of order one decides to stipulate between these discourses, since such relations are clearly not innate. Nobody is born an anthropologist, and, curious though this may seem, still less is anyone born a native. As I stated earlier, anthropology as I understand it begins by asserting the de jure equivalence between the discourses of anthropologist and native, as well as the mutually constitutive condition of these discourses, which only come into existence as such on entering into a relation of knowledge. Anthropological concepts actualize this relation, though this makes them neither true reflections of the native’s culture (the positivist dream), nor illusory projections of the anthropologist’s culture (the constructionist nightmare). What they reflect is a certain relation of intelligibility between the two cultures, while what they project are the two cultures themselves as their imagined presuppositions (as Roy Wagner amply demonstrated). As a result, they perform a double deterritorialization: they amount to transcontextual interfaces whose function is to represent, in the diplomatic sense of the term, the other in the midst of the same, here, there and everywhere. The interminable debate on the universality or otherwise of certain concepts and oppositions therefore seems to me of scant interest. Worse than interminable, this debate is indeterminable: all said and done, everything is relatively universal. The real problem lies in knowing which are the possible relations between our descriptive practices and those employed by other peoples (this is something else Marilyn Strathern taught me). There are undoubtedly many possible relations; but only one impossible relation: the absence of a relation. We cannot learn these other practices — other cultures — in absolute terms; we can only try to make explicit some of our implicit relations with them, that is, apprehend them in relation to our own descriptive practices. Universalizing the Christian metaphysics of body and soul, the modern theory of the social contract or the contemporary biopolitics of kinship is one of the ways of doing just this — of relating. A very unimaginative way, to be sure. But the alternative cannot be the fantasy of an intellectual intuition of other forms of life ‘in their own terms,’ for there is no such thing. ‘Their terms’ are only determined as such in relation to ‘our terms,’ and vice-versa. Every determination is a relation. Nothing is absolutely universal, not because something is relatively particular, but because ‘everything’ is relational. All perfectly obvious, you’ll say. For sure. Admitting the obvious is one thing, though: it’s a very different kettle of fish drawing from it all the possible consequences. In sum, anthropological concepts are relative because they are relational — and they are relational because they are relators. This origin and function is usually marked in the characteristic ‘signature’ of these conceits by a foreign word: mana, totem, kula, potlatch, taboo, gumsa/gumlao… Other no less authentic concepts carry an etymological signature which evokes instead the analogies between the cultural tradition where the discipline emerged and the traditions making up its object: gift, sacrifice, kinship, personhood… Finally, other concepts — equally legitimate — are lexical inventions which seek to generalize conceptual devices of the peoples studied — animism, segmentary opposition, restricted exchange, schismogenesis... — or, inversely, and far more problematically, suck certain more widespread notions from our tradition — the incest prohibition, gender, symbol, culture — into a specific theoretical economy with the aim of universalizing them. It’s clear then that numerous concepts, problems, entities and agents proposed by anthropological theories originate in the imaginative work of the very societies these theories seek to explain. Doesn’t anthropology’s irreducible originality reside in this synergy between the conceptions and practices deriving from the worlds of the ‘subject’ and the ‘object’? Among other plus points, recognizing this would help mitigate our inferiority complex vis-à-vis the ‘natural sciences.’ As Latour observes: The description of kula is on a par with that of the black holes. The complex systems of social alliances are as imaginative as the complex scenarios conceived for the selfish genes. Understanding the theology of Australian Aborigines is as important as charting the great undersea rifts. The Trobriand land tenure system is as interesting a scientific objective as the polar icecap drilling. If we talk about what matters in a definition of science — innovation in the agencies that furnish our world — anthropology might well be close to the top of the disciplinary pecking order. This observation was made, we may recall, in the context of an 1996 AAA-sponsored debate on ‘Science and Anthropology’ Now, the analogy made in this passage is between indigenous conceptions and the objects of so-called natural sciences. This is a possible, and indeed necessary, perspective: anthropology should be able to produce a scientific description of indigenous ideas and practices as if they were objects in the world, or better, in order for them to be objects in the world. (Lest we forget, Latour’s scientific objects are anything but than ‘objective’ and indifferent entities lying patiently in wait of a description.) Another possible strategy involves comparing indigenous conceptions with scientific theories, an approach adopted by Robin Horton, for example, in his ‘similarity thesis.’ Nonetheless, the strategy I advocate here is different again. In my opinion, anthropology has always been somewhat over-obsessed with ‘Science,’ not only in relation to itself — whether it is or isn’t, can or can’t, must or mustn’t be a science — but above all, and this is the real issue, in relation to the conceptions of the peoples it studies: whether to disqualify them as errors, dreams and illusions, and subsequently explain scientifically how and why the ‘others’ fail to produce scientific explanations (of themselves, among other things); or to promote such conceptions as more or less homologous to science, fruits of the same will-to-knowledge driving all humankind: then we end up with Horton’s similarity, or Lévi-Strauss’s science of the concrete. However, the image of science, this gold-standard of thinking, is not the only terrain, nor necessarily the most fertile, for us to relate with the intellectual activity of peoples foreign to the western tradition. If you will allow me a financial metaphor, I’d suggest it’s more interesting for us to float the world conceptual exchange rates, dispensing with the ‘relic of barbarism’ which is mononaturalism, that is, the essentializing reserve currency of a single ontology (to which science enjoys privileged access) capable of guaranteeing the inter-conversion of the various epistemologies. So a different analogy to Latour’s can be imagined. Instead of taking indigenous conceptions as entities akin to black holes or tectonic faults, we can take them as something similar to the cogito or the monad. Paraphrasing our Latour quote, we might say that the Melanesian concept of the person as a ‘dividual’ (M. Strathern) is just as imaginative as the possessive individualism of Locke; that understanding the ‘philosophy of the Indian chieftainship’ (P. Clastres) is just as important as commenting on the Hegelian doctrine of the State; that Maori cosmogony is on an equal par with Eleatic paradoxes or Kantian antinomies (G. Schrempp); that Amazonian perspectivism is just as interesting a philosophical challenge as comprehending the system of Leibniz… Indeed, if it is a question of knowing what matters in evaluating a philosophy — its capacity to create new concepts —, then anthropology, without looking to substitute for philosophy, remains a powerful philosophical tool, capable of airing the stuffy ethnocentric corridors of our philosophy, while freeing us in passing from so-called ‘philosophical anthropology’. In Tim Ingold’s punchy definition: ‘anthropology is philosophy with the people in’. By ‘people’ Ingold intends ‘ordinary people’; but he’s also playing with the meaning of ‘people’ as ‘a people’ or further still as ‘peoples.’ So, a philosophy with other peoples in: the possibility of a philosophical activity which maintains a meaningful relationship with the non-philosophy — the life — of other peoples of the planet, as well as with our own. Not just ‘ordinary people’, therefore, but above all ‘extraordinary’ or ‘uncommon’ peoples, those who live beyond our sphere of ‘communication.’ If real philosophy abounds in imaginary savages, the geophilosophy implied by anthropology strives to articulate an imaginary philosophy with the help of real savages. (In remembrance of Marianne Moore.) I’ve looked at what would happen were we to deny anthropological discourse any epistemological advantage over the native’s discourse. This is the same as asking: what happens when we take native thought seriously? When the anthropologist’s aim ceases to be to explain, interpret, contextualize and rationalize this thought, and becomes one of using it, drawing out its consequences, and ascertaining the effects it may produce on our own? What does it mean to think native thought? Think, I say, without thinking that what we think (the other’s thought) is ‘apparently irrational,’ or, God forbid, essentially rational, but think of it as something remaining unthought within the terms of this alternative — something totally alien to this game? Taking seriously means above all not neutralizing. It means, for instance, bracketing the question of knowing whether and how this thought illustrates cognitive universals of the human species, is a sequel of certain technologies of knowledge transmission, expresses a culturally specific worldview, functionally validates the distribution of political power, and many other forms of neutralizing alien thought. It means suspending this question, or at the very least avoiding enclosing anthropology within it, and taking another tack: deciding, for instance, to think of the other thinking as only (if you will) an actualization of unsuspected virtualities of thought. Everything I’ve just said boils down to the idea that we need to make the notion of symmetrical anthropology reflexive; make it ‘supersymmetrical’, as M. Fukushima once phrased it. But to achieve this aim, it is highly desirable we produce an anthropological concept of the concept, i.e. an anthropological theory of the imagination. As I’ve already spoken way too much, I’ll limit myself here to a few ‘sketchy observations,’ a euphemism, naturally, for ‘peremptory declarations:’ 1. I think it’s about time we rethought the notion of practice. Especially since the radical contrast between theory and practice is, in the end, purely theoretical: pure practice exists only in theory; in practice, it always comes heavily mixed with theory. What I’m trying to say is that the theory of practice, as classically formulated by Bourdieu, supposes a theoretically obsolete concept of theory, which sees the latter as a transcendent meta-practice of a contemplative or reflexive type, existing above and after practice, as its moment of ‘purification’ (in Latour’s sense). In other words, we need a new theory of theory: a generalized theory of theory, one enabling us to think of theoretical activity in radical continuity with practice, that is, as an immanent or constitutive (as opposed to purely regulative) dimension of the intellect embodied in action. This continuity is exactly the same — and this is an important point — as the continuity I identified as obtaining (de jure) in the relation between the discourses of ‘anthropologist’ and ‘native.’ The anthropology of science obviously has a vital contribution to make here, given that one of its core objects is ‘theory in practice:’ the practice of production and circulation of theories. 2. But as a first step we have to resolve our highly ambivalent attitude concerning the propositional model of knowledge. Contemporary anthropology, both in its phenomenological-constructionist and in its cognitive-instructionist guises, has proven notable for insisting on the severe limitations of this model when it comes to dealing with intellectual economies of ‘non-western’ type (I mean non-modern, non-written, non-theoretical, non-doctrinal or non-whatever intellectual economies). Indeed, anthropological discourse has embroiled itself in the paradoxical pastime of heaping propositions on top of propositions arguing for the fundamentally non-propositional nature of other peoples’ discourses — chattering away endlessly about what goes without saying, so to speak. We count ourselves lucky when our natives display a blissful disdain for the practice of self-interpretation, and even less interest in cosmology and system. We’re probably right, since the lack of native interpretation has the great advantage of allowing the proliferation of anthropological interpretations of this lack. Simultaneously, the native’s disinterest in cosmological order fosters the production of neat anthropological cosmologies in which societies are ordered according to their greater or lesser inclination towards systematicity (or doctrinality, or whatever). In sum, the more practical the native, the more theoretical the anthropologist. Let us also not forget that the non-propositional mode is held to be characterized by a constitutive dependency on its ‘context’ of transmission and circulation. This makes it the exact opposite (supposedly, it goes without saying) of scientific discourse — a discourse whose aim is precisely universalization. To repeat a refrain: all of us are context-bound, but some are so much more context-bound than others. My issue here isn’t with the thesis of the quintessential non-propositionality of untamed thought, but with the underlying idea that the proposition is in any sense a good model of conceptuality in general. The proposition continues to serve as the prototype of rational statements and the atom of theoretical discourse. The non-propositional is seen as essentially primitive, as non-conceptual or even anti-conceptual. Naturally, such a state of affairs can be used both ‘for’ and ‘against’ this non-conceptual Other: the absence of rational-propositional concepts may be held to correspond to a super-presence of sensibility, emotion, sociability, intimacy, relational-cum-meaningful engagement in/with the world and what not. For or against, though, all this concedes way too much to the proposition, and reflects a totally archaic concept of the concept, one which continues to define it as the subsumption of the particular by the universal, that is, as essentially a movement towards classification and abstraction. Now, rather than simply divorcing, for better or worse, the concept from ‘cognition in practice’ (to pay homage to Jean Lave’s great book), I believe we need to discover the infra-philosophical, i.e. the vital, within the concept, and likewise (perhaps more importantly) the virtual conceptuality within the infra-philosophical. What kind (or ‘form’) of life, in other words, is virtually projected by ideas such as the Cartesian Cogito or the Kantian synthetic a priori? (Recall Wittgenstein’s indignation against the petty spiritual life presumed by Frazer’s interpretations of primitive rites.) And in like manner, what sort of virtual conceptuality pulsates within Amazonian shamanic narratives, Melanesian initiation rituals, African hunting traps, or Euro-American kinship usages? (Think of the ludicrously stunted conceptual imagination presumed by many an anthropological expatiation upon wild thought.) We need less by way of context and more by way of concept. In other words, we need an anthropological concept of the concept, which assumes the fundamental extra-propositionality of all thought in its integral positivity, and develops in a completely different direction to our traditional notions of ‘innate category,’ ‘collective representation’ and ‘belief.’ In brief, we need an anthropological theory of conceptual imagination: the faculty of creating those intellectual objects and relations which furnish the indefinitely many possible worlds of which humans are capable. This theory must be anthropological, that is, based on the relational matrix of human thinking-and-acting. In Art & Agency, Alfred Gell remarks that anthropological theories must conjoin a theory of social efficacy with cognitive considerations, ‘because cognition and sociality are one.’ Indeed, but the equivalence cuts both ways: a theory of human cognition is relational, i.e. anthropological, or it is nothing. 3. Finally, in order to achieve this we need to draw all the necessary implications from the fact that the native’s discourse speaks about something else besides just the native, that is, his society or mind: it speaks about the world. This means accepting that ‘anthropology’s true problems are not epistemological, but ontological,’ as Vassos Argyrou pithily put it some time ago. And I would like to add: anthropology’s true objects are not epistemologies, but ontologies. I call your attention to the increasingly frequent use of this word, ‘ontology,’ in the contemporary anthropological literature. It strikes me as symptomatic of our growing dissatisfaction with the uncompromisingly Kantian inspiration of our discipline. The image of Being is obviously dangerous analogic ground when it comes to anthropological re-imaginings of non-western conceptual imaginations, and the notion of ontology is not without its own risks. Perhaps Gabriel Tarde’s bold suggestion that we should abandon the irremediably solipsist concept of Being and relaunch metaphysics on the basis of Having (Avoir) — with the latter’s implication of intrinsic transitivity and an originary opening towards an exteriority — is a more enticing prospect in many cases. Nonetheless, I think the language of ontology is important for one specific and, let’s say, tactical reason. It acts as a counter-measure to a derealizing trick frequently played against the native’s thinking, which turns this thought into a kind of sustained phantasy, by reducing it to the dimensions of a form of knowledge or representation, that is, to an ‘epistemology’ or a ‘worldview.’ As if whatever there is to know or view was already decided beforehand — and decided, of course, in favour of our ontology. So the notion of ontology isn’t evoked here to suggest that all thought, be it Greek, Melanesian, African or Amazonian, expresses a metaphysics of Being, but to underline the fact that all thought is inseparable from a reality which corresponds to its exterior. This signifies that the epistemological democracy usually professed by anthropology in propounding the cultural diversity of meanings reveals itself to be, like so many other democracies with which we are familiar, highly relative, since it is based ‘in the final instance’ on an absolute ontological monarchy, where the referential unity of nature is imposed. It is against this pious relativist hypocrisy that I shall conclude by once more claiming that anthropology is the science of the ontological self-determination of the world’s peoples, and that it is thus a political science in the fullest sense, since its motto is — or should be — that which was written on the walls of Paris in May 1968: ‘l’imagination au pouvoir’. The rest is business as usual. Thank you. Acknowledgments Translated (mostly) from the Portuguese by David Rodgers. Some paragraphs have been auto-cannibalized from ‘O nativo relativo’ (Mana, 8/1: 113-148, 2002), and a couple of sentences from ‘Comment on Nurit Bird-David, “Animism Revisited: Personhood, Environment and Relational Epistemology”’ (Current Anthropology, supplement to 40: S67-91). I relied mostly upon the following works while writing the above: Argyrou, V. — 1999. Sameness and the ethnological will to meaning. Current Anthropology, supplement to 40: S29-41. Clastres, P. — 1974. La société contre l’État. Paris: Minuit. Deleuze, G. & F. Guattari — 1991. Qu’est-ce que la philosophie? Paris: Minuit. Fukushima, M. — 2000. On Supersymmetrical Anthropology and an Origin of Science Wars in Japan. (A memo sent by e-mail to EVC on 21/10/2000). Gell, A. — 1998. Art and Agency: An Anthropological Theory. Oxford: University Press. Horton, R. — 1993. Patterns of Thought in Africa and the West: Essays on Magic, Religion and Science. Cambridge: University Press. Ingold, T. — 1992. Editorial. Man 27 (1): 694-697. Latour, B. — 1991. Nous n'avons jamais été modernes. Paris: Ed. La Découverte. — 1996. Not the Question. Anthropology Newsletter 37(3): 1-5. — 2002. War of the Worlds: What About Peace? Chicago: Prickly Paradigm Press. Lave, J. — 1988. Cognition in Practice: Mind, Mathematics and Culture in Everyday Life. Cambridge: University Press. Moore, M. — 1924 1919. Imaginary gardens with real toads in them (‘Poetry’, in Observations, New York: The Dial Press, p. 30-31.) Sahlins, M. — 1995. How ‘Natives’ Think: About Captain Cook, for Example. Chicago: University Press. Scholte, B. — 1984. Reason and Culture: The Universal and the Particular Revisited. American Anthropologist 86(4):960-965. Schrempp, Gregory — 1992. Magical Arrows: The Maori, the Greeks, and the Folklore of the Universe. Madison: University of Wisconsin Press. Stengers, I. — 1997. Cosmopolitiques, vol. 7: Pour en finir avec la tolérance. Paris: Le Découverte & Les Empêcheurs de penser en rond. Strathern, M. — 1988. The Gender of the Gift: Problems with Women and Problems with Society in Melanesia. Berkeley: University of California Press. — 1999. Property, substance and effect: anthropological essays on persons and things. Londres: Athlone. — 1999. No limite de uma certa linguagem: entrevista com Marilyn Strathern. Mana 5(2):157–175. (with C. Fausto & E. Viveiros de Castro). Tarde, G. — 1999 1893. Œuvres de Gabriel Tarde, volume 1: Monadologie et sociologie. Le Plessis-Robinson: Institut Synthélabo. Wagner, R. —1981. The Invention of Culture. Chicago: University Press. 2nd. ed. Martin Holbraad, Anne-Christine Taylor, Bruno Latour, Peter Gow, Marcela Coelho de Souza, Michael Houseman and Marilyn Strathern gave much general inspiration and particular encouragement. Jeanette Edwards, Penny Harvey and Peter Wade were most excellent hosts in Manchester. I thank them all.